In the period before, during, and after surgical treatment, patients are unusually vulnerable to respiratory failure owing to special manifestations of pulmonary edema, atelectasis, hypoventilation, aspiration, sepsis, and hypotension. Awareness of the factors promoting each of these interrelated processes makes possible an effective prevention program or early diagnosis and treatment of perioperative respiratory failure.
Of the forces in the Starling equation governing transcapillary fluid flux, the ones of particular relevance to surgical patients are microvascular hydrostatic pressure and pulmonary capillary permeability.
As indicated, secretion of ADH and aldosterone is a major component of the metabolic response to surgery and trauma.26 Both hormones tend to conserve water and decrease urine output in the postsurgical patient. However, a focus on increasing urine output in the surgical patient by administering large volumes of fluid without regard to this metabolic response could easily result in fluid overload and pulmonary edema, leading to hypoxemia. Guidelines for fluid resuscitation in the perioperative period that focus primarily on urine output and fluid replacements based on empirical values27 can increase extravascular lung water and predispose the surgical patient to perioperative respiratory failure. Although a young, healthy patient with significant cardiopulmonary reserve will undoubtedly tolerate these insults, an elderly surgical patient with a brittle cardiorespiratory status is more likely to develop respiratory failure unless extreme caution is taken with fluid resuscitation, involving close, constant monitoring of central hemodynamics.
Pulmonary edema occurring in the head-injured patient, or neurogenic edema, may be associated with a transient increase in hydrostatic pressure because of intense sympathetic discharge, although it has been suggested that there may be a component of increased capillary permeability as well in this lesion.28 Therefore, monitoring of pulmonary capillary hydrostatic pressure is important in determining therapeutic approaches to the head-injured patient with pulmonary edema.
High-pressure pulmonary edema does not resolve immediately after vascular pressures are normalized.29 The implication of this finding is that the timing of the measurement of pulmonary artery wedge pressure (PAWP), which is used as a reflection of capillary hydrostatic pressure, is crucial in determining whether pulmonary edema is considered to be due to high vascular pressures or to an increase in capillary permeability. Ordinarily, the presence of normal or low PAWP in the presence of pulmonary edema would be regarded as evidence of capillary-leak pulmonary edema. However, if the PAWP is measured during the lag phase of resolution of high-pressure pulmonary edema after PAWP has been decreased, then inappropriate therapy directed at capillary leakage may be given, while the role of the cardiogenic edema that is already resolving slowly is overlooked.
Diuretics such as furosemide clear edema by decreasing the central blood volume and pulmonary capillary hydrostatic pressure.30 However, these agents may produce effects on gas exchange before pulmonary edema has cleared.31 Accordingly, diuretic therapy and fluid management of the oliguric, hypoxemic perioperative patient may confuse the student of critical care at several levels. First, oliguria in the immediate postoperative period is not necessarily due to reduced blood flow to the renal cortex (prerenal oliguria), so fluid challenges aimed at increasing renal blood flow may not be appropriate in this setting. The consequent increase in pulmonary blood volume and pressure predictably increases pulmonary edema. On the other hand, diuretic therapy in such a patient will increase urine output, even when the oliguria is due to reduced renal blood flow, thereby aggravating the prerenal failure. The best approach to this common perioperative conundrum is to recognize that urine output may be an unreliable index of adequate perfusion in the immediate postoperative period, and to seek other indices of perfusion through careful history-taking, physical examination, and first-hand knowledge of the patient's perioperative course. For example, a prior history of congestive heart failure predicts susceptibility to fluid overload, and thus should slow the physician's hand in administering fluid. Similarly, familiarity with the patient's preoperative blood pressure, heart rate, heart sounds, pulse volume, and digital perfusion allow the discerning physician to detect early signs of hypoperfusion requiring volume replacement. Often, the critical distinction between fluid overload and hypovolemia is not clear even to the astute clinician, and in such cases a low threshold for instituting central hemodynamic measurements is indicated.32 Once these steps have been taken, fluid can be administered in the setting of low PAWP, low cardiac output, and oliguria in an attempt to return blood flow toward normal without risking edema. An untoward increase in PAWP without much increase in blood flow warrants immediate re-evaluation of the patient's cardiac function, including an electrocardiogram (ECG) and measurement of cardiac enzymes, and administration of vasoactive drug therapy (e.g., dobutamine, 2 to 10 μg/kg per minute), to increase ventricular contractility and output at reduced PAWP.
Pulmonary Capillary Permeability
A frequent cause of increased capillary permeability and respiratory failure in the surgical patient is unrecognized sepsis, which is commonly seen in the abdomen; this source often requires a surgical or percutaneous radiologic approach. Therefore, a major component of the prevention and treatment of respiratory failure in the surgical patient is the early identification of occult sources of sepsis, aggressive investigation for abdominal causes of sepsis, and the provision of adequate drainage and treatment of septic foci, particularly within the abdomen.
Although both increased microvascular hydrostatic pressure and pulmonary capillary permeability are important factors in the elaboration of extravascular lung water, manipulation of the microvascular pressure (by the use of vasoactive agents and regulation of the state of hydration) is the most direct means of altering pulmonary edema in the surgical patient. A search for a septic focus in the surgical patient is crucial whenever there is evidence of increased capillary permeability. Control of capillary permeability can then be achieved, although only indirectly, by treating the source of sepsis, which may be surgically approachable. The link between sepsis and capillary permeability is thus broken, and the capillary permeability lesion is allowed to resolve with time; its resolution is accompanied by improvement in perioperative respiratory failure. Until the permeability corrects itself, reduction of PAWP to the lowest level associated with adequate peripheral perfusion seems to reduce the edema.30
In the normal lung, ventilation and perfusion are not equally matched, because the shape of the thoracic cavity and the descent of the diaphragm result in greater expansion and ventilation of the lower lobes. Also, blood flow is greater in the dependent areas of the lung during spontaneous ventilation and changes with body position. Therefore the normal lung has an average ventilation:perfusion ratio (/) of approximately 0.8. Many factors in the surgical patient reduce this ratio to very low values, causing hypoxemia, and similar factors lead to resorption of alveolar gas behind closed airways or to compression atelectasis.33,34 This phenomenon of compression atelectasis in the dependent lung is thought to occur within 5 minutes after induction of general anesthesia.
Shunting results from continued perfusion of nonventilated lung units, and the major cause of this imbalance in the surgical patient is perioperative atelectasis, although alveolar edema from fluid overload or capillary leakage could also result in an increase in shunting.
Age, Position, and Airway Closure
Most surgical patients undergo procedures in the supine position, and we are operating increasingly on elderly patients. Also, one of the major effects of surgery is the pain resulting from surgical incisions. Body position, incisional pain, and age all affect the relationship between the functional residual capacity (FRC) and the closing volume. The FRC has been considered the most important index of mechanical abnormality in the lung because it represents the balance of opposing forces on the rib cage at resting lung volume. The closing volume is the measure of gas in the lung at the time when airway closure begins. When FRC exceeds closing volume, lower airway patency is maintained, while airway closure begins when the FRC falls below the closing volume.35 FRC falls with age, and in all patients it is lower in the supine position than in the upright position (Fig. 87-1). The commonly used lithotomy position results in a further decrease in FRC relative to closing volume.
The difference between FRC and closing volume plotted against age in different surgical positions. Both age and position affect airway closure. (Reproduced with permission from Craig DB,Wahba WM, Don H: Airway closure and lung volumes in surgical positions. Can Anaesth Soc J 18:95, 1971.)
When the difference between FRC and closing volume is plotted against the alveolar-arterial oxygen tension gradient (A-a)O2,36 it is evident that the (A-a)O2 oxygen tension gradient increases as FRC falls below closing volume.
Airway closure tends to occur in the most dependent areas of the lung, and in the supine position, more areas of the lung are dependent, thus predisposing the patient to a greater degree of airway closure and hypoxemia. As indicated above,33,34 general anesthesia itself may predispose the patient to compression atelectasis in dependent areas of the lung. Also, in both normal individuals and smokers, increasing age is associated with an increase in closing volume, predisposing the patient to airway closure at higher lung volumes.37 As a group, smokers tend to have higher closing volumes, so that the combination of age and smoking increases the likelihood of significant postoperative hypoxemia. It has generally been accepted that chronic cigarette smoking increases the incidence of postoperative respiratory complications, which may result not only from an alteration in the respiratory defense mechanisms, but also an increase in airway resistance and the work of breathing. It has been demonstrated that cessation of smoking for over 8 weeks is an effective means of decreasing postoperative respiratory complications.38 Although it has been suggested that abstinence too soon prior to surgery may increase the risk of postoperative pulmonary complications, aggressive counseling for smoking cessation prior to any elective surgical procedure still appears to be the best approach.39
Because small airways in the periphery of the lung are not supported by cartilage, they tend to be influenced significantly by changes in pleural pressures. The maintenance of a positive transpulmonary pressure resulting from the negative intrapleural pressure maintains patency of the small airways. Breathing at a reduced FRC, such as occurs with abdominal pain, tends to lead to positive pleural pressures in the dependent areas of the lung, and therefore creates a predisposition to alveolar collapse. Complete collapse results in continued perfusion of nonventilated areas, or shunting; when the airways are merely narrowed, the ventilation:perfusion ratio may be low, which also impairs gas exchange and leads to hypoxemia.
The patient with multiple fractures is at increased risk for developing pulmonary complications, not only from thromboembolic complications, including fat embolism, but also from atelectasis and pneumonia. A major predisposing factor in these patients is the prolonged period of imposed bed rest, particularly in the supine position, with its resultant effect on lung mechanics and lung volumes. Early operative stabilization of fractures in these patients has been shown to decrease pulmonary morbidity40 because it allows more effective respiratory physiotherapy and early ambulation, as well as frequent changes in body position to minimize dependant alveolar volume loss.
A major cause of morbidity in traumatic quadriplegic patients is respiratory failure secondary to loss of use of the intercostal muscles of respiration.41 It has been suggested that the best position for respiratory therapy in these patients is from horizontal to 35° head-up,42 whereas the maximum FRC is achieved in the 60° to 90° head-up position.
Upper Abdominal Surgery and Diaphragm Dysfunction
Although many of the factors discussed above are present in patients undergoing most surgical procedures, the most serious sequelae are found in patients undergoing upper abdominal procedures. In these patients there is a significant fall in vital capacity (VC) almost immediately postoperatively, within the first 4 hours.43 There is a slower but definite fall in FRC, which peaks at about 24 hours and is associated with significant hypoxemia. In most patients with no pre-existing lung disease, this effect of upper abdominal intervention on VC and FRC does not result in clinically significant respiratory complications. However, in patients who already have abnormalities of gas exchange, these effects can lead to severe respiratory failure. The postoperative decrease in VC is primarily a restrictive rather than obstructive phenomenon, as evidenced by the maintenance of a normal ratio between the forced expiratory volume at 1 second and the forced vital capacity (FEV1/FVC).44 This restriction may be related to incisional pain, which decreases the patient's ability to cough and clear secretions, and eventually leads to an increase in closing volume and a decrease in FRC. If not corrected, a fall in VC results in atelectasis and hypoxemia and a decrease in FRC. To correct this abnormality, transcutaneous electrical nerve stimulation has been used to provide postoperative analgesia after abdominal surgery.45 Epidural analgesia and intercostal blockade have also been used for this purpose. Although all of these techniques have produced improvements in VC and FRC, none immediately returns VC or FRC to preoperative values. This suggests either that these techniques do not adequately control pain, or that pain is not the only cause of postoperative respiratory dysfunction after upper abdominal surgery.
Patients undergoing upper abdominal operations have a significant decrease in the maximal transdiaphragmatic pressure at FRC, which is not altered by use of epidural analgesia.46 This finding suggests that the respiratory dysfunction after upper abdominal surgery may result from a primary effect of the procedure on diaphragmatic function. Ford and coworkers showed that there is a switch from predominantly abdominal breathing to rib cage breathing in the postoperative period in patients undergoing upper abdominal surgery (Fig. 87-2).47 Diaphragmatic dysfunction was similarly identified in an animal model undergoing cholecystectomy.48 These studies suggest that general anesthesia may not be responsible for the postoperative diaphragmatic dysfunction. Mere traction on the gallbladder in an animal model also produced similar effects on diaphragmatic function.49
Relationship between the ratio of abdominal to rib cage diameter and time after abdominal surgery. Interrupted lines represent individual patients and the solid line represents the mean values for these four patients. Note the switch from predominantly abdominal breathing preoperatively to rib cage breathing postoperatively. (Reproduced with permission from Ford et al.47)
Although open cholecystectomy has been associated with significant depression in postoperative pulmonary function, several reports50–52 have demonstrated less impairment of postoperative pulmonary function following laparoscopic cholecystectomy. There still is a decrease in FRC immediately after the operation, but it is much smaller and of significantly shorter duration than with the open procedure, and the VC and FRC return to essentially preoperative levels within 24 hours.52 Therefore from the respiratory standpoint, laparoscopic cholecystectomy is superior to open cholecystectomy and should be the preferred method for critically ill patients requiring this procedure. The increase in intra-abdominal pressure with pneumoperitoneum associated with the laparoscopic procedure has a minimal hemodynamic effect, and in patients with decreased cardiopulmonary reserve may prove significant, warranting close hemodynamic monitoring in the operating room in such patients.53 As indicated in Chap. 89, less aggressive procedures such as percutaneous drainage of the biliary tract may be indicated in situations in which the patient is too unstable to be taken to the operating room or to be subjected to a general anesthetic.
Apart from the factors identified above, aging has been associated with reduced elastic lung recoil, decreased expiratory flow rate, and diminished airway protective reflexes.37 Obesity is also a major risk factor for postoperative pulmonary complications, because these patients tend to breathe at reduced lung volumes, so closing volume frequently exceeds FRC, leading to hypoxemia and atelectasis.53a The increased work of breathing produced by the increased mass also contributes to respiratory dysfunction. Not only the type of operation, but the location of the incision tends to affect the degree of respiratory impairment seen in the postoperative period.54 In open cholecystectomy, the subcostal incision tends to produce less impairment than a midline incision. The severity of postoperative lung impairment decreases in the following order: thoracic surgery, upper abdominal surgery, lower abdominal surgery, and superficial surgery (Fig. 87-3).
Postoperative changes in VC for different surgical incisions. Note that the upper abdominal surgery patients have the greatest postoperative depression in VC. (Reproduced with permission from Ali et al.43)
As pointed out in earlier chapters, shock and pulmonary edema in the form of cardiac failure also affect diaphragmatic function through changes in diaphragmatic force as well as glycogen depletion in diaphragmatic muscle.55 Table 87-1 summarizes preventive measures as well as some of the factors that reduce the FRC and increase closing volume in the postoperative patient.
Table 87–1. Perioperative Atelectasis ||Download (.pdf)
Table 87–1. Perioperative Atelectasis
|Component of the Tendency Toward Atelectasis||Promoting Factors||Preventing Factors|
|Reduced functional residual capacity||Supine position||45° upright position|
|Ascites||Positive end-expiratory pressure|
|Upper abdominal incision||Analgesia|
|Increased closing volume||Age||Preoperative physiotherapy|
|History of smoking||Smoking cessation|
|Airway secretions||Cough, suction, deep breathing|
|Pulmonary edema||Avoidance of overhydration|
In surgical patients, hypoventilation is characteristically caused by impairment of ventilation resulting from the restrictive effect of painful incisions or peritonitis. It may also result from central nervous system (CNS) depression due to anesthesia, analgesia, or CNS injury. The increased metabolic requirement after injury places a significant demand on the respiratory system. When calories, particularly in the form of carbohydrates, are provided to match this increased energy expenditure, the increase in CO2 production necessitates a significant increase in ventilation to maintain normocapnia.56 In surgical patients with significant pulmonary reserve, this added demand can be met without untoward effects. However, in depleted surgical patients with borderline respiratory reserve, this extra demand may precipitate respiratory failure or lead to other manifestations such as prolonged ventilator dependency.
The respiratory system is protected from sepsis and atelectasis by a respiratory control mechanism that responds to hypoxemia, hypercapnia, acidosis, and the presence of irritating or noxious stimuli in the airway. These mechanisms can be significantly depressed in the postoperative patient as a result of anesthesia or excessive narcotic analgesia. Inhalational anesthetics are known for their respiratory depressive effect, which results in alveolar hypoventilation and a reduced response to carbon dioxide, as well as a blunted response to hypoxemia and acidosis.57 In the postoperative period, narcotic analgesics may have undesirable effects. Whereas in optimal doses they decrease abdominal pain and increase the ability to cough and clear secretions, in larger doses they may depress the respiratory center, producing alveolar hypoventilation as manifested by hypercapnia and secondary hypoxemia.
The cough reflex is the main mechanism by which particles are cleared from the upper airway. The cough response is altered not only by anesthesia, but also by narcotic agents. Clearance of particles from the lower airways depends primarily on the mucociliary system, which can be disturbed by several factors in the postoperative period. Anesthetics alter ciliary activity and mucus production which leads to the production of mucus plugs that block the lower airways. In addition, the cellular defense mechanisms of the respiratory system may be altered by anesthetic agents.58
The supine position and depression of normal protective reflexes during general anesthesia predispose the surgical patient to aspiration of gastric acid, which is one of the major causes of perioperative morbidity and mortality.59 This event can first produce airway obstruction (from aspirated debris and chemically induced bronchoconstriction), then a chemical burn of the airway (with fluid loss into the injured area), an intense inflammatory response, and finally lung infection. The clinical presentation of patients with gastric acid aspiration varies widely. Very mild cases present only with transient coughing and minimal bronchospasm; the most severe cases exhibit a progressive downhill course characterized by hypovolemia, hypoxemia, and finally fulminant bacterial pneumonia.
The treatment of acid aspiration consists of the following: (1) rapid removal of debris by immediate suction (endotracheal intubation and fiberoptic bronchoscopy may be necessary at this stage); (2) placement of a nasogastric tube to evacuate the stomach and prevent further episodes; (3) oxygen administration and mechanical ventilation if indicated by the degree of respiratory failure; (4) bronchodilator therapy if bronchospasm is significant; (5) maintenance of normovolemia and normal perfusion by monitoring and replacement of lost fluid, as well as vasoactive and inotropic support where necessary; and (6) treatment of pneumonia by appropriate antimicrobial agents on the basis of Gram stain and culture findings. Steroids have not been of any benefit in treating these patients. Preventive measures that can be taken in high-risk patients to prevent the aspiration of low-pH gastric contents include gastric decompression and the use of agents that decrease gastric acid production, although there is evidence that this may predispose to nosocomial pneumonia.60–62
Predicting and Preventing Perioperative Lung Dysfunction
Many attempts have been made to correct the postoperative abnormalities in lung function, using techniques such as incentive spirometry, intermittent positive-pressure breathing (IPPB), and nasal continuous positive airway pressure (CPAP).63–65 Although incentive spirometry has been reported to be ineffective in decreasing postoperative pulmonary complications following cardiac and upper abdominal surgery,64 IPPB, incentive spirometry, CPAP,65 and physiotherapy generally improve postoperative respiratory function; IPPB offers no advantage over physiotherapy when the latter is maximized in the postoperative period.66 Although nonyielding abdominal binders have a further restrictive effect on lung volumes postoperatively, the elastic binders may produce some benefit.67 It must be recognized, however, that none of these methods completely reverses the postoperative respiratory dysfunction.
Attempts have been made to predict postoperative pulmonary morbidity by assessing respiratory mechanics preoperatively, as well as by identifying risk factors such as age, obesity, smoking, and location of incisions. No individual respiratory parameter predicts respiratory morbidity or mortality in an individual patient. In general, however, the poorer the preoperative respiratory function, the more likely the patient is to have severe postoperative respiratory complications. Based on cumulative experience, the following spirometric criteria for predicting morbidity and mortality in postoperative adult patients have been proposed.68,69 If the FEV1 is <1 L, the FVC <1.5 L, the FEV1/FVC is <30%, or the forced expiratory flow [FEF25% –75%] is <0.6 L/s, and if the maximum minute ventilation is <50% of the predicted value, then the risk of postoperative pulmonary complications is very high. In patients whose respiratory function is below this threshold, the strategy is to provide treatment that will improve respiratory function to a level above this threshold. Such treatment may involve cessation of smoking, diaphragm muscle conditioning, weight loss, and the treatment of heart failure, fluid overload, and any identifiable reactive airway disease.
A patient who undergoes lung resection is at even greater risk of postoperative pulmonary complication, particularly if the excised lung tissue was functional.68,69 In these patients, the effect of the lost lung volume must be considered along with the factors discussed above. A quantitative perfusion lung scan can help to predict the postoperative pulmonary spirometric performance of these patients by indicating how much of the lung will remain after the planned procedure. The postoperative FEV1 is then calculated as the product of the preoperative FEV1 and the fractional perfusion of the remaining lung. The usual rule for an adult patient is that the operative risk is prohibitive if the predicted postoperative FEV1 is ≤0.8 L. The prediction can be made more accurate by also measuring the diffusing capacity, which is an independent predictor of morbidity and mortality after major lung resection. A useful guideline is to exclude from major lung resection all patients whose diffusing capacity is <60% of the predicted value, even if spirometric values are considered satisfactory.69 Patients with only slightly impaired pulmonary function (FEV1 and diffusing capacity ≥80% predicted) with no cardiovascular risk factors can undergo pulmonary resections including pneumonectomy without further investigation. For others, exercise testing as well as pulmonary split-function test studies are recommended. The symptom limited cardiopulmonary exercise testing measures the maximum volume of oxygen utilization (VO2max) as an index of pulmonary and cardiovascular reserve. A VO2max <10 mL/kg per minute is generally considered a contraindication to any resection, whereas a value >20 mL/kg per minute or >75% of predicted normal is considered safe for major resections. Resections that involve no more than one lobe usually lead to early functional deficit followed by recovery, and permanent loss in pulmonary function is usually <10%. Generally, pulmonary function tests tend to overestimate the functional loss after lung resection.70 Arterial blood gas criteria may also be used to exclude patients from major lung resection because of the prohibitive risk of postoperative morbidity and mortality. Patients who have a room-air partial pressure of arterial oxygen (PaO2) of <50 mm Hg or a partial pressure of carbon dioxide (PCO2) of >45 mm Hg at rest are considered to have a prohibitive operative risk and should not undergo major pulmonary resection. Other forms of surgical intervention are justifiable in the presence of these blood gas criteria only if they are considered mandatory and lifesaving.
Treatment Principles for Perioperative Respiratory Failure
At present, no specific therapy exists for underlying diaphragmatic dysfunction. Therefore the principles of respiratory care in the surgical patient are as follows:
Maximization of the preoperative respiratory status. Meeting this goal may entail cessation of smoking, diaphragmatic conditioning exercises, reduction in obesity, and treatment of any identified cardiorespiratory disease, including congestive heart failure, bronchopneumonia, or bronchospasm.
Aggressive physiotherapy and early ambulation to overcome the effects of the supine position on changes in lung volumes, particularly the relationship between closing volume and FRC. In patients with multiple fractures, early operative stabilization will decrease the period of recumbency.
Adequate treatment of sepsis and shock, with recognition of the important role of surgery or interventional radiology in the identification and drainage of areas of sepsis such as intra-abdominal abscesses.
Judicious use of intravenous fluids to maintain adequate perfusion while avoiding overhydration and pulmonary edema.
Optimal use of analgesics to control pain without producing respiratory depression.
Administration of supplemental oxygen to hypoxemic patients to improve arterial oxygenation.
Treatment (as specific as possible) of any identifiable cause of hypoxemia; for example, bronchodilator therapy for a patient with bronchospasm, or antibiotic therapy directed against a specific organism isolated in a patient with a pneumonic process.
Preoperative pulmonary assessment, especially in patients with poor pulmonary reserve and most especially in those undergoing lung resection.This will allow an assessment of the relative risk of postoperative morbidity.
Institution of mechanical ventilation when the above measures fail. Frequently, mechanical ventilation can be avoided if strict attention is paid to preventive measures. Also, vigorous application of these principles immediately after the patient is stabilized can shorten the duration of mechanical ventilation considerably.71